Upgrade to v1.3

This guide covers upgrading Orchestrator deployments from v1.2 to v1.3. Upgrades which skip versions (e.g. v1.1 to v1.3) are not explicitly documented or supported at this time.

The v1.3 upgrade follows a standard procedure:

• Upgrade the Terraform modules
• Bump version numbers
• Run data migrations

Prerequisites

There are quite a few data migrations to run after this upgrade. We recommend taking a DB snapshot of your RDS instances before continuing so that there is a path to downgrade.

Build and publish the application containers on the head of the release branch by following the documentation, and package and upload version 1.4.36 of the orc8r Helm chart as well.

If you are using local Terraform state (the default), ensure all Terraform state files (i.e. terraform.tfstate) are located in your working directory before proceeding. This means terraform show should list existing state (rather than outputting No state).

Upgrade Terraform Modules

Set the source values for each of the Orchestrator modules in your root Terraform module to point to this release's modules and bump chart and container versions:

module orc8r {
  source = "github.com/magma/magma//orc8r/cloud/deploy/terraform/orc8r-aws?ref=v1.3"
  # ...
}

module orc8r-app {
  source = "github.com/magma/magma//orc8r/cloud/deploy/terraform/orc8r-helm-aws?ref=v1.3"
  # ...
  orc8r_chart_version = "1.4.36"
  orc8r_tag           = "MAGMA_TAG"  # from build step, e.g. v1.3.0
}

You should bump your chart version from 1.4.35 to 1.4.36 and orc8r_tag to whatever semver tag you published your new Orchestrator container images as.

IMPORTANT: check the "Build Orchestrator" section for the Git hash to check out before packaging the orc8r Helm chart. You must deploy 1.4.36, not 1.4.37 on the head of the release branch.

Then,

terraform init --upgrade
terraform refresh

Terraform Apply

Before upgrading the deployment, manually scale the Prometheus pods to 0 replicas, otherwise the Prometheus version upgrade in this release will fail:

kubectl --namespace orc8r scale deployment orc8r-prometheus --replicas=0

terraform apply     # Check this output VERY carefully!

After applying Terraform changes, all your application components should be upgraded and Prometheus should be back up on v2.20.1.

Data Migrations

NOTE: If you're upgrading to release tag v1.3.0 specifically, m014_enodeb_config is not required.

$ export CNTLR_POD=$(kubectl --namespace orc8r get pod -l app.kubernetes.io/component=controller -o jsonpath='{.items[0].metadata.name}')
$ kubectl --namespace orc8r exec -it ${CNTLR_POD} -- bash

(pod)$ cd /var/opt/magma/bin

(pod)$ ./m013_relay_split

...

(pod)$ ./m013_policy_ipv6

...

(pod)$ ./m014_enodeb_config

...